INDEX

Explanations

symbols and keywords related to programming or code structure

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.38

 Inter

-0.34

po

-0.33

 final

-0.33

reg

-0.32

Far

-0.32

ignac

-0.32

POSITIVE LOGITS

:+:

0.91

AddTagHelper

0.78

IntoConstraints

0.73

niſſe

0.66

 ***!

0.65

Personensuche

0.63

########.

0.63

<unused79>

0.62

<unused28>

0.62

<pad>

0.62

Activations Density 0.008%

symbols and keywords related to programming or code structure

No Comments

No Known Activations